Self-censorship in Singapore: what young people say they will not say

There is a particular silence in Singaporean political conversation that is easy to misdescribe. It is not that young people have no views, and it is not, for most of them, that they expect to be prosecuted for holding them.

It is that saying something publicly has costs they can name, and the costs are not hypothetical.

What the calculation actually contains

Asked directly, the reasons cluster into four, and only one of them is about the state.

Employment. A post is searchable, permanent and attached to a full name. Hiring in a small labour market involves people who know people. The concern is not a formal blacklist; it is that a recruiter, a manager or a client will read something and form a view, and that nothing will ever be said about why an application went nowhere.

Family. In a country this size, a public position reaches relatives quickly. For many the immediate deterrent is not the government but an uncle at the next family gathering, and an argument that will be repeated at every subsequent one.

Permanence. What is written at twenty is retrievable at forty. Young people are unusually aware of this, having watched it happen to others, and they discount their own future views accordingly.

The law. Defamation, contempt, the correction powers under the Protection from Online Falsehoods and Manipulation Act, and the rules around public assembly. This is real and it is fourth on the list, not first.

Why the order matters

If the deterrent were principally legal, the remedy would be legal. Change the law and speech follows.

Because the deterrent is mostly social and economic, the law could change tomorrow and the silence would largely persist — because the recruiter, the uncle and the permanence of the internet would all still be there.

That is the uncomfortable part of the finding, and it is why “Singaporeans are afraid to speak” misdescribes the situation. Most are not afraid in the way that phrase implies. They have made a calculation about cost and benefit in which the benefit of speaking publicly is low and diffuse, and the cost is specific and lands on them.

The part that is about the law

The fourth reason deserves its own treatment, because it is not nothing.

Defamation actions in Singapore have been brought by public figures against critics, and the judgments have been substantial. The correction powers under Pofma require a person to carry a notice stating that what they published contains a falsehood, and although the obligation is to correct rather than to remove, in practice almost everyone who received a direction removed the post.

Neither of those is a criminal sanction for an opinion. Both establish that publishing something about a public matter can be expensive in ways that do not apply to publishing about anything else, and that is sufficient to shape behaviour without any prosecution occurring.

The practical effect is a shift toward the safest available formulation. People who would have made a claim make an observation instead; people who would have named a decision-maker describe a decision. The content survives, and the specificity does not.

Where the conversation goes instead

The silence is public rather than total, and this is the most consistent finding in any account of it.

Political conversation among young Singaporeans happens in private messaging groups, in closed chats, among friends, at the dinner table. It is often blunt, well-informed and considerably less cautious than anything the same people write in public.

That produces a distorted public record. A reader sampling only what is posted openly would conclude that a generation holds mild, uncommitted views. What is actually happening is a sorting: the views are held, and the channel has been selected for its audience.

What changes it

Three things visibly do, and none of them is exhortation.

Anonymity. Pseudonymous participation removes the employment and family costs at once, which is why anonymous forums carry political argument of a different temperature from the same people’s named accounts.

Professional cover. People whose work makes commentary normal — academics, journalists, lawyers, those working for civil society organisations — speak more freely because the cost structure is different. Speaking is part of the job rather than a risk to it.

Collective settings. A position stated as one of many at a forum, a dialogue or a class carries less individual exposure than the same position posted alone under a full name.

The common element is that each one lowers the personal cost rather than raising the benefit. Nothing makes speaking more rewarding; what works is making it cheaper.

Why this is worth measuring rather than deploring

There is a familiar and unhelpful version of this discussion in which young Singaporeans are described as apathetic, and a second in which they are described as oppressed. Neither survives contact with what people actually say about their own reasoning.

What they describe is a cost-benefit judgement made by people who are informed, who care about the subject, and who have concluded that the specific act of publishing under their own name is not worth what it costs them. That is a rational response to a set of conditions, and it tells you about the conditions rather than about the people.

The conditions are also not fixed. Employment norms change, platforms change, and the size of the consequence attached to a public position is not a constant. Whether the calculation comes out differently for the next cohort is an empirical question, and the answer will be visible in what they post rather than in what anyone says about them.

The classroom problem

Institutions that teach politics run into this directly, and their experience is a useful test of the account above.

Students who will not post under their own names will argue at length in a seminar room, will write sharply in assessed work, and will interview politicians without difficulty. The same person is articulate in one setting and silent in another, which rules out the explanations that appeal to disposition — shyness, apathy, lack of interest.

What differs between the settings is audience and permanence. A seminar has a bounded audience and leaves no searchable record. A post has an unbounded audience and leaves a permanent one.

That is the whole of the difference, and it is why interventions aimed at confidence do not work. Nobody involved lacks confidence.

What the silence costs

Two costs follow, and they are borne by different people.

The first is borne by the person. A view that is never stated publicly is never tested publicly, and the correction that comes from being argued with by a stranger does not happen. Private-channel discussion tends to occur among people who already agree.

The second is borne by everyone else. Public discussion that systematically omits a demographic’s views describes a country that does not exist, and decisions get made against that description. Surveys partly compensate, but a survey captures a position without the reasoning behind it, and the reasoning is the part that would change anyone’s mind.

Neither cost is dramatic and both are cumulative, which is the hardest kind of cost to act on.

The generational claim, examined

It is commonly asserted that younger Singaporeans are more willing to speak than their parents were, and that the trend continues.

The evidence for it is mixed. Younger cohorts are visibly more willing to hold positions that diverge from official framings, and to say so in private. Whether they are more willing to say so publicly, under their own names, in a form an employer can find, is a different question — and the permanence of the internet is a deterrent their parents never faced.

It is possible for a generation to be simultaneously more heterodox and more careful. On the accounts young people give of their own reasoning, that is the combination Singapore currently has.

What this is not

Two things are worth ruling out, because both are commonly asserted and neither matches what people describe.

It is not a claim that Singapore is uniquely silent. Every labour market in which employers search candidates’ names produces some version of this calculation, and the permanence of the internet is not a local condition. What varies is the size of the consequence and the size of the society in which it lands, and Singapore is small enough that both weigh more.

It is also not a claim that nobody speaks. People do — under their own names, at cost, and some of them pay it. The account above describes a distribution rather than an absolute, and the interesting question is about the shape of the distribution rather than about whether it has a tail.

A note on measuring this

Any account of self-censorship has a methodological problem built into it: the people most affected are the least likely to describe it publicly.

Surveys reach them, and surveys of this subject consistently report higher caution than the visible public discussion would suggest. But a survey asking whether someone self-censors is asking them to report on a decision they may not have experienced as a decision at all.

Where to check this

The legal provisions referred to are in the Protection from Online Falsehoods and Manipulation Act, the Public Order Act, and the law of defamation as applied by the Singapore courts. Parliamentary debates on the Pofma legislation, including the arguments made about its effect on public discussion, are in Hansard, the official verbatim record of proceedings published by the Parliament of Singapore. Correction directions issued under the Act, and the corresponding clarifications, are published on the government’s Factually website.